[Palaeontology, Vol. 49, Part 2, 2006, pp. 405–420] VARIATIONS IN THE MORPHOLOGY OF EMU (DROMAIUS NOVAEHOLLANDIAE)TRACKS REFLECTING DIFFERENCES IN WALKING PATTERN AND SUBSTRATE CONSISTENCY: ICHNOTAXONOMIC IMPLICATIONS by JESPER MILA` N Geological Institute, University of Copenhagen, Øster Voldgade 10, 1350 Copenhagen K, Denmark; e-mail: [email protected] Typescript received 23 October 2003; accepted in revised form 9 March 2005 Abstract: Fossil footprints appear in a variety of preserva- walls collapse, destroying the shape of the footprint. tional states, each revealing a different morphology that can Increased speed of progression affects the shape of the foot- give rise to misidentification and misinterpretations. Com- print dramatically as the distal ends of the digits become parative ichnological work was conducted using living emus more deeply impressed in the substrate during acceleration. (Dromaius novaehollandiae). It was clearly demonstrated Plantigrade stance adopted by the emu while feeding produ- that the morphological variation that occurred in footprints ces highly elongated footprints. Applying these observations of the same animal, walking in the same manner, was to the study of fossil footprints demonstrates that great care caused by variation in substrate consistency. Dry sand sub- should be paid to the original sedimentary conditions at strates are unlikely to preserve any anatomical details of the the time of track making, as well as to the stance and gait foot, whereas damp sand or mud of firm consistency pre- of the trackmaker. serves a high level of anatomical detail. The finest anato- mical details, such as skin impressions, are only preserved Key words: Emu tracks, field experiments, sediment pro- in firm mud or clay. In semi-fluid to fluid mud the track perties, footprint morphology. That variation in substrate consistency exercises a a track. If the sediment is saturated with water, sand will strong control on the morphology of fossil tracks and be too loose to preserve tracks and clay will be too fluid. traces has long been well documented among marine The optimal parameters for track preservation, according invertebrate trace fossils (Bromley 1996). Only recently to Laporte and Behrensmeyer (1980), are when the sedi- has the same awareness about the relationship between ment is moist, and has a grain size between sand and the morphology of a vertebrate track and the consistency clay. of the substrate in which it is emplaced been the topic of Allen (1997) made several observations on subfossil systematic study. Bromley (1996, fig. 7.1) used his own human and cattle tracks in the Severn Estuary, south-west footprint emplaced on photographic paper, dry sand, England. A human footprint imprinted in deep, fluid to damp sand and in deep mud on an intertidal mud flat to semi-fluid mud would collapse and the sediment would demonstrate how different tracks from the same action gradually flow back and fill the track, obscuring it from can appear in different substrates. the bottom up. Such a track would, if fossilized, reveal lit- Laporte and Behrensmeyer (1980) described the con- tle about the nature of the trackmaker, and may be recog- nection between grain size and water content of the sedi- nized only as a mass of disrupted sediments below a ment and the potential for tracks to be preserved, on the slight depression in the sediment surface (Allen 1997). basis of field observations of tracks in Recent and Tracks emplaced in soft mud have a much less pro- Plio ⁄ Pleistocene sediments in Kenya. According to their nounced tendency to flow and collapse, but are generally observations, tracks are unlikely to be preserved in dry poor in detail because of the tendency for the mud to sediments, since dry sand is too loose to preserve tracks, adhere to the trackmaker’s foot and create adhesion and dry clay is simply too hard to allow the formation of spikes as the foot is withdrawn, leaving the footprint ª The Palaeontological Association 405 406 PALAEONTOLOGY, VOLUME 49 blurred. In some cases the movement of the foot leaves flightless birds to use for comparison with theropods, striations on the track walls. Tracks preserved in such because of its didactyl foot. To help interpret strange col- sediments are likely to show only gross anatomical fea- lapsed theropod tracks from the Upper Triassic deposits tures (Allen 1997). Firm mud, according to Allen’s (1997) of Jameson Land, East Greenland, Gatesy et al. (1999) observations, is likely to produce very well-defined but used a turkey, Meleagris gallopavo, and a helmeted guinea- shallow tracks and, if fossilized, should preserve even fine fowl, Numida meleagris, running and walking in mud of details, such as skin impressions. These observations were different consistencies, producing tracks at several states of backed up by an intensive study of cross-sections through collapse. Recent studies by Farlow et al. (2000) and Smith artificially produced tracks in layered plasticine, to reveal and Farlow (2003) of the interspecific variations in foot- and describe the various subsediment deformation struc- prints and foot morphology of the ratites and other curso- tures that occur beneath and around a vertebrate foot- rial birds further demonstrate the value of incorporating print. A similar study, designed to describe the influence studies of extant animals in palaeoichnological studies. that substrate consistency exercises on the morphology of The aim of this paper is to describe the range of tracks and undertracks was carried out by Mila`n and morphological variation in tridactyl footprints, owing to Bromley (in press) using vertical sections through emu differences in substrate consistency and mode of progres- tracks emplaced in layered cement packages of different sion, by using an emu as a trackmaker. The extant tracks consistencies. are described and comparisons made with fossil tracks Diedrich (2002) demonstrated several different preser- and trackways that show similar sediment-induced differ- vational variants of Triassic rhynchosaurid tracks caused ences in morphology. by differences in water content of the sediments. Tracks made in dry subaerial sediments consist of little more than faint claw imprints. With increasing water content of the METHODS sediment, shallow tracks are found having skin texture preserved. In more water-rich and thus softer sediments, The emus used for the experiment belong to breeder the tracks become deeper and more blurred in shape, and Karin Holst, Mønge, Denmark. In order to record tracks, finally, subaquatic tracks produced by a swimming track- two types of sediment were used as substrate: (1) local maker are found as elongated, parallel scratch traces. organic-rich, dark soil from within the emu enclosure Using live animals has several advantages over previous mixed with different quantities of water, in order to experiments using models of feet. While the track experi- record tracks in mud having consistencies from firm to ments using artificial model feet (Allen 1997; Manning liquid; and (2) glaciofluviatile sand from a nearby sand- 2004) are easier to conduct and, importantly, much easier pit. Each sediment type was analysed following Tucker to document as one is in total control of all parameters, (2001), with the mean grain-size, median grain size and one important factor is missing: the dynamic interaction degree of sorting calculated by the methods advised by between the animal and the substrate emphasized by Folk and Ward (1957). The local organic-rich soil is Baird (1957). By using live animals, all the factors result- poorly sorted (degree of sorting, 1Æ57), and has a median ing from differences in gait, individual behaviour and grain size of 2Æ2 and a mean grain size of 2Æ18 mm. The mode of progression are reflected in the footprints. glaciofluvatile sand is poorly sorted (degree of sorting, The extant emu, Dromaius novaehollandiae, and other 1Æ42), and has a median grain size of 2Æ4 and mean grain large cursorial birds are the best living analogues to size 2Æ38 mm. The sand grains are angular with a high Mesozoic theropods. The emu and the rhea, Rhea ameri- sphericity in the terminology of Tucker (2001). cana, having a pedal skeleton and footprint morphology Eight different substrates were prepared in which the resembling that of non-avian theropods, are especially emus were encouraged to walk: dry loose sand, damp sand, obvious candidates for comparative ichnological work. wet sand, thin layer of soft mud, deep firm mud, deep The first comparison of footprints from ratitous birds semi-firm mud, deep semi-fluid mud and deep fluid mud. and theropods was by Sollas (1879), who compared casts In order to obtain an emu track with as many anatomical of emu and cassowary tracks with what he then believed details preserved as possible, a fresh, severed emu foot was to be tracks of giant extinct birds in the Triassic con- impressed in a sheet of soft potter’s clay. This track serves glomerates of South Wales. Padian and Olsen (1989) as a reference for discussing the amount of preserved ana- demonstrated that the stance and gait of theropods and tomical details in the tracks from the field experiments. small bipedal ornithopods were similar to those of the To record longer trackways, sand patches were laid out extant rhea, by comparing trackways from rheas with on the paths preferred by the emus within the fenced those of theropods. Farlow (1989) made similar observa- paddock. Selected tracks and trackways were recorded tions of tracks and trackways of an ostrich, and pointed photographically and relevant measurements were taken in out that an ostrich might not be the best of the large the field. Where the consistency of the sediment allowed, MILA` N: VARIATIONS IN MORPHOLOGY OF EMU TRACKS 407 plaster casts were made. The plaster casts depicted herein emu. Digit III is the longest with the shorter digits II and are curated by the Geological Museum, University of IV of subequal length, making the foot close to symmetri- Copenhagen, numbers MGUH 27474–27479.